Training Generative Verbal Behavior in

1
Training Generative Verbal Behavior in Early
Intensive Behavioral Intervention
John McElwee Step-By-Step Ian Stewart National
University of Ireland, Galway
2
Training GVB in EIBI
OVERVIEW OF THE WORKSHOP

• Early Intensive Behavioral Intervention
• Generative Verbal Behavior
• Relational Frame Theory
• EIBI Critical features of RFT
• The ABLA
• The ARPSP
• The ABLLS
• Advanced Training

3
EIBI

• EIBI
• What is it?
• Brief History
• Critique Generative Verbal Behavior

4
Early Intensive Behavioral Intervention
Q. What is it?
A. An effective BA-based approach to
remediation of deficits for ASD children, with
language as the main focus
(See, e.g., Harris and Weiss, 1998)
5
Early Intensive Behavioral Intervention
Treatment begins as soon as the child is
diagnosed with a Pervasive Developmental Disorder
typically between 2 to 3 years
(See, e.g., Smith, 1999 Sallows, 2005)
6
Early Intensive Behavioral Intervention
Initially, one-to-one treatment for 30 to 40
hours, extended over approximately 18 months with
subsequent social skills development and school
enrolment
(See, e.g., Smith, 1999)
7
Early Intensive Behavioral Intervention
Treatment is guided by the application of basic
principles of behavior (Michael, 1993 Matson et
al. 1996) and a curriculum of skills from basic
to complex (e.g., Lovaas, 1981)
8
EIBI

• EIBI
• What is it?
• Brief History
• Critique Generative Verbal Behavior

9

• Initial phase

• UCLA Young Autism Project
• 1987 Lovaas Study
• Maurice (1993)
• Let Me Hear Your Voice
• Shook (1993)
• Behavior Analysis Certification Board
• Maurice, Green Luce (1996)
• Behavior Intervention for Young Children with
  Autism
• Leaf McEachin (1999)
• Work Progress

Early Intensive Behavioral Intervention
10

• Paradigm Expansion

• Skinner (1957)
• Verbal Behavior
• Sundberg Partington (1998)
• Teaching Language to Children
• Sundberg Michael (2001)
• The Benefits of Skinners Analysis of Verbal
  Behavior
• Pelios Lund (2001)
• A Selective Overview of Issues on Classification,
  Causation and EIBI for Autism
• Carbone (2004)
• A Selective Overview of Issues in EIBI

Early Intensive Behavioral Intervention
11
Verbal Behavior Classification
Controlling Variable
Verbal Relation
UMO/CMO
Yes
Mand
No
Non-Verbal Stimulus
Yes
Tact
No
Verbal Stimulus with Point-to-Point Match
No
Intra-verbal
Yes
Echoic / Imitation / Copying Text
With Formal Similarity
Without Formal Similarity
Textual / Transcriptive
12
EIBI CURRICULA
13
Early Learning Curriculum
Pairing with Positive Reinforcement Manding Food
Drink Toys Activities Vocal Signing
Pictures Visual Performance Matching Objects and
Pictures Imitation Gross Fine Oral
Motor Echoics Sounds Blends Words Receptive Si
mple motor action Points to
objects Tacts Labels reinforcers Objects
Pictures Intraverbals Fill in blanks
14
EIBI

• EIBI
• What is it?
• Brief History
• Critique Generative Verbal Behavior

15
Critique of the EIBI approach

• Although EIBI works for approximately 40 of
  children, for whom mainstreaming is achieved,
  there is still a substantial number for whom
  progress is slow
• Children learn specific and robot-like
  performances and cant shift between ideas
• This appears to be a problem of GENERATIVE VERBAL
  BEHAVIOR

16
Generative Verbal Behavior

• Generative Verbal Behavior
• What exactly is GVB?
• How do we approach it?

17
GENERATIVE VERBAL BEHAVIOR

• Richard Malott
• How can we understand a sentence we've never
  heard before and how can we say a meaningful
  sentence we've never heard or said before?

Malott, R.W. (2003). Behavior analysis and
linguistic productivity. Analysis of Verbal
Behavior, 19, 11-18.
18
GVB in human language
What is a jumjaw?
Is a jumjaw the same as a cat?
Is a jumjaw bigger than a tractor?
Does a jumjaw have ears?
Which one of these is NOT a jumjaw?
Lassie, Old Smoky, Katrina, Tweety
If I said there was a jumjaw at the back of the
room, how would you react?
19
GVB in human language
A jumjaw is a dog
Is a jumjaw the same as a cat?
Is a jumjaw bigger than a tractor?
Does a jumjaw have ears?
Which one of these is NOT a jumjaw?
Lassie, Old Smoky, Katrina, Tweety
If I said there was a jumjaw at the back of the
room, how would you react?
Do you own a jumjaw?
Exercise Make up a sentence with the word jumjaw.
20
BA and GVB

• This problem of training GVB is not confined to
  EIBI but is a problem in ABA more generally
• The ABA approach to language is based on Skinners
  Verbal Behavior which has been criticized as
  not providing a theoretical or empirical basis
  for GVB (e.g., Chomsky, 1959)
• So how do we approach GVB?

21
Generative Verbal Behavior

• Generative Verbal Behavior
• What exactly is GVB?
• How do we approach it?

22
Q. How do we explain this phenomenon from a
behavioural perspective?
Q. How do we train it?
Q. How do we assess it?
23
What GVB is NOT
It is NOT stimulus generalization
DOG
?
JUMJAW
24
What GVB is NOT
It is NOT direct contingency training
A DOG IS A JUMJAW
Where is the contingency training?
JUMJAW
25
What GVB is NOT
It is NOT response induction
DOG
WHAT IS THIS?
JUMJAW
WHAT IS THIS?
?!
26
What GVB is NOT
It is NOT Pavlovian conditioning
1
2
WHAT WAS THAT I SAW EARLIER?
LATER ON
3
4
LATER STILL
JUMJAW
THAT WAS A JUMJAW
27
So where do we start?
28
STIMULUS EQUIVALENCE SIDMAN (1971)
Trained
Tested
Spoken word (A)
Written word (C)
Picture (B)
29
Trained Relations
Untrained / Emergent / Derived Relations

• See notes

30
Critique of the EIBI approach

• LACK OF LANGUAGE PRODUCTIVITY
• Children learn specific performances
• Robot-like performances
• Cant shift between ideas

Stimulus equivalence / Derived relations
Why does stimulus equivalence occur?
Where does the productivity of both language and
stimulus equivalence come from?
31
Relational Frame Theory
Q. What is it?
A. A Behavior Analytic Account of Human
Language Cognition

• See notes

32
Q.
Why does stimulus equivalence occur?
Where does the productivity of both language and
stimulus equivalence come from?
Relational Frame Theory
A.
Both language and stimulus equivalence are
examples of relating behavior
RFT explains relating as learned behavior
33
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

34
Relational Frame Theory
Theoretical Background
35

• Sidman was using conditional discriminations when
  he first demonstrated the phenomenon of stimulus
  equivalence
• So lets start by looking at the conditional
  discrimination

• See notes

36
Conditional Discrimination
)
Conditional Stimulus


U
Discriminative Stimulus (for as Conditional
Stimulus)
Discriminative Stimulus (for as Conditional
Stimulus)

• See notes

37
Conditional Discrimination Emergent Performances
TRAIN CONDITIONAL DISCRIMINATIONS

U


)
)
)

• SYMMETRY

• TRANSITIVITY

TEST FOR DERIVED / EMERGENT PERFORMANCES

• EQUIVALENCE

U





U


)

• See notes

38
Directly Trained Baseline Relations

• See notes

Derived Symmetrical Relations
Derived Transitive Relations
Derived Equivalence Relations
39
Trained Relations
Untrained / Emergent / Derived Relations
40
Q. How does RFT explain emergent responding such
as stimulus equivalence and language?
A. As arbitrarily applicable relational
responding
41
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

42
What is Relational Responding?
CONDITIONAL DISCRIMINATION
Conditional Stimulus
IDENTITY MATCHING
Discriminative Stimulus (when SC green triangle)
Discriminative Stimulus (when SC red circle)
43
Generalised Identity Matching
Training Exemplars
Generalization
44
Non-Arbitrary Relational Responding

• Same
• (Identity)

• Different
• (Oddity)

• Bigger

• Smaller

(Comparison)
45
Non-Arbitrary Relational Responding
Physical Relations
Arbitrarily Applicable Relational Responding

???
CONTEXTUAL CUE
46
Arbitrarily Applicable Relational Responding

CONTEXTUAL CUE
JUMJAW
IS SAME AS
JUMJAW
DOG
???
DOG CAT PIG
CONTEXTUAL CUE
47
Normally developing English speaking adults
respond in accordance with a wide variety of
contextual cues for arbitrarily applicable
relations (AAR)
is the same as

is equivalent to
means
is equal to
is
For example, here are some of the many cues for
just the AAR of sameness alone
parallels
is similar to
is like
is identical to
48
GENERATIVITY in human language
PSYCHOLOGICAL EXERCISE
BEFORE
AFTER
(A jumjaw is a dog)
(Its a nice day)

• How many of you own
• a jumjaw?

• How many of you own
• a jumjaw?

• Do you like jumjaws?

• Do you like jumjaws?

• What do you like
• best about jumjaws?

• What do you like
• best about jumjaws?

CONTEXTUAL CUE
Readily answer questions
Little or no reaction
49
Stick in list of questions asked before
is a
DOG
is a
Jumjaw
Any question you could previously answer with
respect to Dog can now be answered with
respect to Jumjaw
50
Summary So Far Arbitrary and Non-Arbitrary
Relations
From the perspective of RELATIONAL FRAME THEORY
there is an important distinction to be made
NON-ARBITRARY (PHYSICAL) RELATIONS
ARBITRARILY APPLICABLE RELATIONS
A
B
SAME
IS
W
Z
MORE THAN
Wolf
Dog
BIGGER THAN

• SUMMARY SO FAR see notes

51
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

52
Learning Arbitrarily Applicable Relations

• The fundamental conception is as follows
• All animals that demonstrate complex forms of
  learning are capable of learning to respond to
  non-arbitrary relations between stimuli
  (e.g., smaller than, lighter than)
• However, language-able humans also have a
  prolonged history of learning to respond to the
  relations among stimuli where the relations are
  not defined by the relata but by additional
  contextual cues
• Learning to name objects is perhaps one of the
  earliest and most fundamental forms of
  arbitrarily applicable relational responding

53
Learning Arbitrary Relational Responding
When children are taught to name objects, they
are provided with masses of explicit
bi-directional training
CRel
Contextual cue for Name-Object / Object-Name
responding

• See notes

54
Further Examples
Learning to name a toy teddy bear
PARENT
CHILD
PARENT
Where is Teddy?
There (points to toy)
Good boy!
What is this? (holding toy)
Teddy
Good boy!
Learning to name a fruit
PARENT
CHILD
PARENT
Where is the apple?
There (points to apple)
Good boy!
What is this? (holding toy)
Apple
Good boy!
55
After thousands of such interactions, the child
no longer requires explicit bi-directional
training
Another example
PARENT
CHILD
PARENT
There (points at novel object)
Where is the peach?
Good boy!
What is this? (holding peach)
Peach
No Consequence
56
Origins of Arbitrary Relational Responding

• In other words, the masses of bi-directional
  training involved in the parental naming game
  allow the child to derive the untaught
  symmetrical relation from the object peach to the
  name peach without being explicitly trained to do
  so
• In effect, the word is, and the naming context
  more generally, bring the relational frame of
  sameness (also called co-ordination) to bear on
  the name peach and the object peach

57
Stimulus equivalence is also an example of
responding in accordance with the relational
frame of sameness, but now with three or more
stimuli as relata
58
Origins of Arbitrary Relational Responding

• Aspects of the context that control responding in
  the case of stimulus equivalence are the
  matching-to-sample context itself
• Consider, for example, how many early education
  exercises involve asking a child which picture
  is the same as or goes with a written word

59
Example 1 An Early Education Exercise Using
Matching-To-Sample
CRel
60
Example 2 An Early Education Exercise Using
Matching-To-Sample
CRel
TREE
DOG
HOUSE
61

• In effect, the matching-to-sample format alone
  may be able to invoke sameness or equivalence
  responding in a child with the appropriate
  history of explicit bi-directional and explicit
  equivalence training

62
Furthermore, during a childs early years he or
she is also explicitly taught about equivalence
relations (note that this learning, like learning
to name, often occurs through informal verbal
interactions)
63
Example 2 Learning Equivalence Through Language
Interactions
Explicitly Train
Television
SAME
Telly
SAME
TV
TV
SAME
Television
After many more explicitly trained examples of
sameness, the child can derive relations without
explicit training
Train
Dog
SAME
Mutt
SAME
Hound
Test
Hound
SAME
Dog
64
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

65
Types of Relational Frame

• Relational Frame Theory asserts that there are a
  number of relational frames other than sameness
• Furthermore, the derived performances these
  frames involve vary widely
• Relational Frames include
• Same
• Opposite
• Different
• Comparison (More / Less)
• Hierarchy
• Perspective

66
Non-Arbitrary Relational Responding
Physical Relations
NB Change below to language
Arbitrarily Applicable Relational Responding

???
CONTEXTUAL CUE
67
SAME
If A is the same as B then B is the same as A if
A is the same as B and B is the same as C, then A
is the same as C and C is the same as A
B
A
dog
A B gt B A
B C gt C B

• See notes

Mutual Entailment
C
dog
A B B C gt A C
Combinatorial Entailment
68
Normally developing English speaking adults
respond in accordance with a wide variety of
contextual cues for arbitrarily applicable
relations (AAR)
is the same as

is equivalent to
means
is equal to
is
For example, here are some of the many cues for
just the AAR of sameness alone
parallels
is similar to
is like
is identical to
69
SAME (EQUIVALENCE)
B
A
dog
Transitivity
Symmetry
COMBINATORIAL ENTAILMENT
C
MUTUAL ENTAILMENT
dog
Taught
Untaught
70
Non-Arbitrary Relational Responding
Physical Relations

• Bigger Than

• Smaller Than

Arbitrarily Applicable Relational Responding
gt

• Bigger Than

• Smaller Than

lt
???
CONTEXTUAL CUES
71
MORE / LESS (COMPARISON)
If A is more than B then B is less than A if A
is more than B and B is more than C, then A is
more than C and C is less than A
gt
B
A
lt
gt
A gt B gt B lt A
gt
lt
B gt C gt C lt B

• See notes

lt
Mutual Entailment
C
A gt B B gt C gt A gt C
Combinatorial Entailment
72
Arbitrarily Applicable Relational Responding
MORE THAN
Which one would you take to the shop?
73
Arbitrarily Applicable Relational Responding
Non Arbitrary Relation

• 5 fen

More Than

• 5 jiao

Arbitrarily Applicable Relation

• 5 fen

0.1
Less Than

• 5 jiao

74
OPPOSITE
The abstract relation of oppositeness has the
property that an opposite of an opposite is the
same, an opposite of an opposite of an opposite
is an opposite, and so on
B
A
hot
cold
Opposite to
Same as
C
Taught
Untaught

• See notes

75
Example of Training OPPOSITE during language
learning
Near
OPPOSITE
Far
OPPOSITE
Close
Close
SAME
Near
After many more explicitly trained examples
of opposite, the child can derive the same
relation from the two opposite relations
Train
Pretty
OPP
Ugly
OPP
Beautiful
Test
Beautiful
SAME
Pretty
76
DISTINCTION / DIFFERENCE
The relational frame of distinction, or
difference, derives yet other relations. For
example, if A is different to B, then B is
different to A. However, if A is different to B
and B is different to C, then the relation
between A and C remains unspecified (i.e., A and
C could be the same or different).
B
A
Car
House
?

• Mutual
• Combinatorial

C
e.g. 1, Money
Taught
e.g. 2, Auto
Untaught
77
Normally developing English speaking adults
respond in accordance with numerous contextual
cues for arbitrarily applicable relations (AAR)
same as
different from
before / after
in front of / behind
bigger / smaller than
right / left of
there is a multiplicity of AAR
type / class of
analogous to
above / below
opposite to
78
Further frames include . . .

• See notes

79
Further frames include . . .
Perspective-Taking
I
You
Here
There
Now
Then

• Explain more extensively

• See notes

80
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

81
Transformation of Functions
The property that gives arbitrarily applicable
relational responding its psychological
significance, is the Transformation of Stimulus
Functions
If a stimulus has a particular psychological
function (e.g., Fear) then in a context that
selects that function as appropriate, the
functions of related stimuli may also be
transformed in accordance with the relation

• See notes

82
Defining the word Function

• In EIBI, function means an action associated
  with an object
• e.g., Function of a knife cut

• In Relational Frame Theory function is defined
  broadly to mean any elicited or evoked response
  on the part of the organism
• e.g.,

• An alarm clock may evoke getting up

• An dog may elicit anxiety

83
Transformation of Functions
If someone has a fear of dogs, and they are told
that jumjaw is another word for dog, then fear
may be elicited upon hearing Here comes a
jumjaw!
Psychological Function
FEAR
is a
DOG
is a
Psychological Function
Jumjaw
FEAR

• See notes

84

• See notes

Limoo
same as
Betrang
Betrang
Limoo
same as
Jumjaw
Wogget
Which one is yellow?
Transformation of Functions

• Function Consume
• Feature Yellow
• Class Fruit

85
is
Betrang
Lemon
SAME
SAME
What is the look of a Betrang?
What is the taste of a Betrang?
TASTE
LOOK
Sour
Yellow
Bumpy

• See notes

86
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

87
RELATIONAL FRAME THEORY A SUMMARY

• Defining properties of Relational Frames
• Multiple Exemplar Training
• Contextual cues Crel Cfunc
• RFT An operant account

88
RELATIONAL FRAME THEORY A SUMMARY
Defining Properties of Relational Frames
1. Mutual Entailment
3. Transformation of Functions
2. Combinatorial Entailment
89
RELATIONAL FRAME THEORY A SUMMARY
Multiple Exemplar Training
Explicitly trained
CRel (e.g. is)


CRel (e.g. is)
Sr
tree
tree
Sr
90
RELATIONAL FRAME THEORY A SUMMARY
Multiple Exemplar Training
CRel (e.g. smaller than)
Trained
Derived
91
RELATIONAL FRAME THEORY A SUMMARY
Crel Cfunc
Crel (e.g. smaller than)
predicts
Crel (e.g. bigger than)
Cfunc (e.g. worth)
92
RELATIONAL FRAME THEORY A SUMMARY
RFT An Operant Account
Explicitly Reinforced Relational Response
Multiple Exemplars
Explicitly Reinforced Relational Response
Unreinforced Derived Relational Response
93
RFT

• RFT
• Background
• Relational Responding
• Non Arbitrary
• Arbitrarily Applicable Relational Responding
  (AARR)
• Origins of AARR
• Relational Frames
• Transformation of Function
• Summary
• RFT Training Protocols

94
RFT Training Protocols

• Lipkens, Hayes Hayes (1993)
• Combining RFT Skinnerian VB
• Training Derived Manding

95
Testing Derived Relational Responding in a
Normally Developing Infant

• Based on empirical evidence (e.g., Lipkens, Hayes
  Hayes, 1993 ) children seem to develop a
  receptive derived relational repertoire before
  they show an expressive derived relational
  repertoire and will develop derived or verbal
  exclusion after they have shown naming

Lipkens, G., Hayes, S.C. Hayes, L.J. (1993).
Longitudinal study of the development of derived
relations in an infant. Journal of Experimental
Child Psychology, 56, 201-239.
96
Lipkens, Hayes Hayes (1993)

• Pretraining (16 mths)
  (Test of auditory comprehension)

• Where is the teddy?

• Where is the rabbit?

16 months
97
Lipkens, Hayes Hayes (1993)

• Picture-Name Training

• This is TAK, can you say TAK?

• This is OS, can you say OS?

• Name-Picture Testing

Receptive Mutually Entailed Relational Responding
17 months

• Where is TAK (or OS)?

98
Lipkens, Hayes Hayes (1993)

• Name-Picture Training

• Where is OEF?

• Where is UI?

• Picture-Name Testing

• What is this?

• What is this?

Expressive Mutually Entailed Relational Responding
19 months
99
Lipkens, Hayes Hayes (1993)

• Picture-Name Training

• This is TAK

• This is OS

• Picture-Sound Training

• This goes WOOH

• This goes PAT-PAT

• Name-Sound Testing

• What does TAK say?

• What does OS say?

Combinatorially Entailed Relational Responding
23 months
100
Lipkens, Hayes Hayes (1993)

• Picture-Name Test

• Where is the DOG?

• Picture-Name Exclusion Test

Exclusion (Non-Verbal)
16 months

• Where is NOVEL NAME?

101
Lipkens, Hayes Hayes (1993)

• Picture-Name Test

• Where is the DOG?

• Picture-Name Exclusion Test

• Where is NOVEL NAME?

• What is this?

Verbal Exclusion
27 months
102
Multiple Exemplar Training Mutual Entailment

• Picture-Name / Name-Picture Training

• This is __. Can you say__ ? (P-N)

• Where is __ ? There is __. (N-P)

Need practice doing this exemplars But they
also need practice at the basic skills
103
Multiple Exemplar Training Mutual Entailment

• Name-Picture / Picture-Name Training

• Where is __ ? This is __. (N-P)

• What is this? This is __ . (P-N)

104
Multiple Exemplar Training Combinatorial
Entailment

• Picture-Name Training

• This is NAME

A

• Picture-Sound Training

• This goes SOUND

B
C

• Name-Sound Training

• NAME goes SOUND SOUND is what NAME goes

105
RFT Training Protocols

• Lipkens, Hayes Hayes (1993)
• Combining RFT Skinnerian VB
• Training Derived Manding

106
Note Protocols Combining Relational Frame Theory
With Skinners Verbal Behavior

• RFT distinguishes between two different types of
  each of Skinner's (1957) verbal operants
• Non-Verbal (Based on Direct-Acting Contingencies)
• Cannot be distinguished readily from any other
  form of social behavior
• Do not possess any of the referential or symbolic
  qualities usually ascribed to verbal events
• Verbal (Based on Arbitrarily Applicable
  Relational Responding)
• Can be distinguished from other forms of social
  behavior
• Appear to possess the symbolic qualities ascribed
  to human language

107
The Tact
Echoic Response is Emitted
Echoic Reinforced as Mand
Correct!
Rabbit
1
2

• The tact is a discriminative response in which
  the presence of a particular object is
  discriminative for the emission of a name
  response
• Production of the correct response produces
  reinforcement

108
Tact
Dog
1
The Verbal Tact
Learning an arbitrary relation of co-ordination
2
3
Verbal Tact
Hound
Hound means Dog

• A particular tact (e.g., Dog) is part of the
  childs repertoire
• The child learns that Dog and Hound are in a
  relation of co-ordination
• The child can demonstrate a verbal tact

109

• Picture-Name Training

• This is TAK, can you say TAK?

• This is OS, can you say OS?

• Name-Picture Testing

Receptive Mutually Entailed Relational Responding

• Where is TAK (or OS)?

• Tacting is itself a vital part of the repertoire
  required for the development of both receptive
  arbitrarily applicable relational responding

110

• Name-Picture Training

• Where is OEF?

• Where is UI?

• Picture-Name Testing

• What is this?

• What is this?

• as well as expressive arbitrarily applicable
  relational responding

111
The Nonverbal Mand
Echoic Response is Emitted
Echoic Reinforced as Mand
Juice
1
2

• The echoic is emitted as a non-arbitrary
  relational response
• At some point a particular echoic is reinforced
  with the reinforcer which it specifies (for the
  caregiver)

112
The Nonverbal Mand
Mand Response is Emitted
Mand Reinforced as Mand
Juice, Juice, Juice . . .
3
4

• The reinforcement of a particular echoic such as
  Juice with actual juice (the mand) now emerges
• This and other mands are examples of
  discriminative responses in which the emission of
  the mand in a particular context leads to
  reinforcement

113
The Verbal Mand
Get Johnny a drink -- there's some juice in the
icebox
Juice and Drink now participate in a frame of
coordination
Verbal Mand is Emitted
Verbal Mand is Reinforced

• The verbal mand develops at a later stage, when
  the child has developed a repertoire of
  arbitrarily applicable relational responding and
  particular mand responses are framed as in
  co-ordination or other types of relation with
  other verbal responses (e.g., in this example,
  drink and juice are equivalent)

Drink
114
The Verbal Mand
Verbal Mand is Emitted
The Relational Network is Tested
Drink
There's no juice -- what about something else?
Drink and pear participate in a frame of
difference
The Relational Network is Reinforced
Pear
115
RFT Training Protocols

• Lipkens, Hayes Hayes (1993)
• Combining RFT Skinnerian VB
• Training Derived Manding

116
Training Derived Manding (CABAS, 2002)

• 3 normally-developing children 3 children with
  a diagnosis of autism (categorised as
  speaker/early readers) attending a CABAS school
  in Dublin were participants
• A desk and three chairs were used in all
  experimental phases, with participants seated
  facing Experimenter No.1
• Experimenter No. 1 conducted all tests, while
  Experimenter No. 2 functioned as a second data
  recorder (for inter-rater reliability)

117
Experimental Sequence

• Conditional Discrimination Training
  (A1-B1, B1-C1, A2-B2, B2-C2)
• Train and Test for a Derived Transfer of Mand
  Functions (A1-C1, A2-C2)
• Reversed Conditional Discrimination Training
  (B1-C2, B2-C1)
• Test for a Reversed Derived Transfer of Mand
  Functions (A1-C2, A2-C1)
• Re-training Baseline Conditional Discriminations
  (B1-C1, B2-C2)
• Test for a Derived Transfer in accordance with
  Re-trained Baseline Conditional Discriminations
  (A1-C1, A2-C2)

118
(No Transcript)
119
A1
B1
C1
Yellow
Golden
Transfer of Mand function
Mand function
120
Derived Manding

• During mand training a participant manded for
  tokens by handing experimenter an A1 card for a
  pink token and an A2 card for a yellow token
  corrective feedback is presented
• The participant was required to mand for the
  appropriate tokens with NO errors across the
  three mand responses on each trial
• During mand testing a participant manded for
  tokens by handing the experimenter a C1 card for
  a pink token and a C2 card for a yellow token
  the C stimuli did NOT appear on the box lids and
  NO corrective feedback was presented

121
Results

• All 6 children showed a derived transfer of mand
  functions and 4 showed ABA reversals in derived
  manding
• The children diagnosed as autistic showed greater
  difficulty in mastering the performances than the
  normally developing children
• One child required exemplar training and
  demonstrated a gradual improvement in derived
  manding across exemplars
• But reversals demonstrated control over derived
  manding, and the blind tester controlled for
  cuing
• The study showed the utility of combining the
  concepts of Verbal Behavior with RFT

122
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

123
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

124
NON ARBITRARY
ARBITRARY
Ball
125
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

126
Specific
Generalized
127
Relational Frame Theory
Operant Response
Non Arbitrary Relational Responding
"Correct!"
A
B
C
128
Relational Frame Theory
Operant Response
Arbitrarily Applicable Relational Responding
"Correct!"
A
B
C
A
B
C
A
B
C
129
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

130
CORE CONCEPTS OF RFT
MULTIPLE EXEMPLAR TRAINING
CONTEXTUAL CONTROL
131
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

132
CORE CONCEPTS OF RFT
MULTIPLE EXEMPLAR TRAINING
CONTEXTUAL CONTROL
133
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

134
CORE CONCEPTS OF RFT
MULTIPLE EXEMPLAR TRAINING
CONTEXTUAL CONTROL
135
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

136
Training Implications
Simple Discrimination
EASY
HARD
Successive
Simultaneous
137
(No Transcript)
138
Critical Features of RFT

• Critical Features of RFT
• Two Core Concepts
• Arbitrary vs Non Arbitrary Relations
• Specific vs. Generalized
• The RFT Matrix
• Two core concepts
• Multiple Exemplar Training
• Contextual Control
• Simultaneous / Successive
• Giving this workshop

139
GIVING THIS WORKSHOP
EIBI
RFT
ABLA, ABLLS and other EIBI METHODOLOGIES
WHOLLY NEW RFT-INSPIRED CURRCIULUM?
RFT IMPROVEMENTS TO ABLA, ABLLS and other EIBI
METHODOLOGIES?
SOME OF BOTH
140
Predicting Outcome of EIBIWeiss et al., 2006,
Sallows et al., 2006

• Core curriculum targets and the Early Learning
  Measure (Weiss, M.J. 1999 Smith,T. et al., 2000)
• Matching to Sample
• Imitation of Actions
• Imitation with Objects
• Verbal Imitation
• Receptive Commands/Actions
• Tacting/Labeling

141
Results

• Rate of learning core skills is highly correlated
  with best outcome results
• IQ gains and classroom placement
• Correlations in high 80 range
• Pretreatment presence of core skills (especially
  imitation) highly predictive of best outcome

142
ABLA

• ABLA
• Introducing the ABLA
• ABLA Research
• RFT Conceptualization of the ABLA

143
ASSESSMENT OF BASIC LEARNING ABILITIES (ABLA)
144
Assessment of Basic Learning Abilities
(ABLA) (Kerr et al., 1977)
Level 1 Imitation A tester puts an object into a
container and asks the client to do likewise
Level 2 Position Discrimination When a red box
and a yellow can are presented in a fix position,
a client is required to consistently place a
piece of green foam in the container on the left
when the tester says "put in."
Level 3 Visual Discrimination When a red box and
a yellow can are randomly presented in left-right
positions, a client is required to consistently
place a piece of green foam in the yellow can
when the tester says "Put it in."
Level 4 Match to Sample Discrimination When
allowed to view a yellow can and a red box in
randomly alternating left-right positions, and is
presented randomly with a yellow cylinder and a
red cube, the client consistently places the
yellow cylinder in the in the yellow can and the
red cube in the red box.
Level 5 Auditory Discrimination When presented
with a yellow can and a red box (in fixed
positions) a client is required to consistently
place a piece of foam in the appropriately
container when the tester randomly says "red box"
(in a high pitched rapid fashion) or yellow can
(in a low pitched drawn-out fashion).
Level 6 Auditory-Visual Discrimination The same
as level 5 except that the right-left positions
of the container is randomly alternated.
145
ABLA Level 1
Stimulus array Experimenter puts gray foam into
container
Required response Child required to put gray
foam into container
146
ABLA Level 2
Stimulus array Red Box and Yellow Container in
fixed position across trials Gray foam as object
Required response Child must put gray foam into
container on the left
147
ABLA Level 3
Stimulus array Red Box and Yellow Container vary
in position across trials Gray foam as object
Required response Child must put gray foam into
the can
148
ABLA Level 4
Stimulus array Red Box and Yellow Container vary
in position across trials Red cube and yellow
cylinder as objects
Required response Child must put red cube into
red box and yellow cylinder into yellow can
149
ABLA Level 5
RED BOX
YELLOW CAN
Required response Child must put gray foam into
the red box when experimenter says red box and
must put it into the yellow can when experimenter
says yellow can
Stimulus array Red box and yellow can in fixed
positions across trials Gray foam as object
Experimenter says either Red box or Yellow can
150
ABLA Level 6
RED BOX
YELLOW CAN
Required response Child must put gray foam into
the red box when experimenter says red box and
must put it into the yellow can when experimenter
says yellow can
Stimulus array Red box and yellow can vary in
position across trials Gray foam as object
Experimenter says either Red box or Yellow can
151
ABLA

• ABLA
• Introducing the ABLA
• ABLA Research
• RFT Conceptualization of the ABLA

152
Recent research on additional discriminations to
add to the ABLA
Level Visual-Visual Nonidentity Matching The
same containers are used as the other ABLA tasks.
The sample stimuli consisted of a silver colored
piece that was shaped into capital letters
spelling the word BOX and a purple piece of wood
that shaped into the uppercase and lowercase word
Can. The client was required to place the wooden
object into the receptacle that corresponded to
the wooden word. (Sako et al., 2004)
Level Auditory-Auditory Identity Matching A
tester says a word (e.g., "pen"), one assistant
says the same word while another assistant says a
different word. The client points to the
assistant who said the same word (Marion et al.,
2003)
Level Auditory-Auditory Nonidentity Matching A
tester says "ball" on some trials and "ice" on
other trials. Two assistants say either "rink" or
"field". The client must learn to point to the
assistant who say" rink" when he hears ice and
the assistant who says" field" after he hears
ball. (Marion et al., 2003)
153
Generalizations from research on the ABLA (Martin
and Yu 2000)
Test levels are hierarchically ordered in
difficulty
Failed levels are difficult to teach
The ABLA has high predicative validity for
performance on other tasks
The ABLA is a better predictor of a client's
performance than experienced staff with direct
knowledge of that client
The ABLA has shown concurrent validity with
language and reading
Mismatch of the ABLA test level of clients to
ABLA difficulty of training tasks causes aberrant
behaviors
Direct-care staff with no knowledge of the ABLA
test often mismatch clients to difficulty level
of training task
154
ABLA

• ABLA
• Introducing the ABLA
• ABLA Research
• RFT Conceptualization of the ABLA

155
RFT Conceptualization of ABLA
Level 1 Simple Responding
Level 2 3 Visual Discrim. (position as extra
dimension in 2 but not 3)
NON-RELATIONAL
Level 4 Non Arbitrary Visual Cond. Discrim.
NON-ARBITRARY RELATIONAL
AAIM
SAME MODALITY
VVNM, AANM
Level 5 6 Arbitrary A-V Cond. Discrim.
(position as extra dimension in 2 but not 3)
ARBITRARY RELATIONAL
AAIM, AANM
DIFFERENT MODALITY
156
ARPSP

• ARPSP
• Introducing the ARPSP
• Examples of ARPSP stages
• Ongoing Development of the ARPSP
• Future Directions

157
Pre-Requisite Training
TRAINING DIMENSIONS
LEVEL 1 SIMPLE DISCRIMINATION
MODALITY Visual, Auditory, Tactile, Olfactory,
Kinesthetic
Simultaneous Successive Visual
Simultaneous Successive Auditory
Sound No Sound Sound White Noise Sound -
Sound Sound Word Word Syllable Word - Word
NO. OF COMPARISONS 2, 3, 4
LEVEL 2 NON-ARBITRARY CONDITIONAL
DISCRIMINATION
Multiple Dimensions
SIMULTANEOUS vs SUCCESSIVE
Assessment of Relational Precursors Skill
Progression (ARPSP)
LEVEL 3 ARBITRARY CONDITIONAL DISCRIMINATION
Contextual Control
CATEGORIES Super-ordinate vs subordinate
Multiple Dimensions
LEVEL 4 MUTUAL ENTAILMENT
SINGLE vs CROSS MODALITY
Multiple Dimensions
LEVEL 5 COMB. ENTAILMENT
Multiple Dimensions
SPECIFIC ? GENERALIZED
158
ARPSP

• ARPSP
• Introducing the ARPSP
• Examples of ARPSP stages
• Future Directions

159
Computer Based Intervention
1 NON ARBITRARY AUDITORY IDENTITY MATCHING
2 NON ARBITRARY VISUAL IDENTITY MATCHING
3 CONTEXTUALLY CONTROLLED NON ARBITRARY AUDITORY
IDENTITY MATCHING
4 CONTEXTUALLY CONTROLLED NON ARBITRARY VISUAL
IDENTITY MATCHING
5 AUDITORY TO VISUAL MTS WITH AUDITORY NAME AS
THE SAMPLE
6 MUTUALLY ENTAILED SOUND OBJECT / OBJECT SOUND
RELATIONS (I.E., TRAIN SOUND-OBJECT TEST
OBJECT-SOUND)
7 TWO PICTURE COMBINATIONS AS SAMPLE WITH S/D
CUES AS COMPARISONS
8 PICTURE AUDITORY STIMULI COMBINATION AS
SAMPLE AND S/D CUES AS COMPARISONS
9 COMBINATORIAL ENTAILMENT TRAINING / TESTING
160
1 NON ARBITRARY AUDITORY IDENTITY MATCHING
161
VERY GOOD!
162
2 NON ARBITRARY VISUAL IDENTITY MATCHING
163
3 CONTEXTUALLY CONTROLLED NON ARBITRARY AUDITORY
IDENTITY MATCHING
164
4 CONTEXTUALLY CONTROLLED NON ARBITRARY VISUAL
IDENTITY MATCHING
165
5 AUDITORY TO VISUAL MTS WITH AUDITORY NAME AS
THE SAMPLE
166
6 MUTUALLY-ENTAILED SOUND OBJECT / OBJECT SOUND
RELATIONS (I.E., TRAIN SOUND-OBJECT TEST
OBJECT-SOUND)
167
6 MUTUALLY-ENTAILED SOUND OBJECT / OBJECT SOUND
RELATIONS (I.E., TRAIN SOUND-OBJECT TEST
OBJECT-SOUND)
168
ARPSP

• ARPSP
• Introducing the ARPSP
• Examples of ARPSP stages
• Future Directions

169
Insert slide on future dirctions of research for
ARPSP here
E.G., EMPIRICAL TESTS, TRAINING
170
ABLLS

• ABLLS
• Introducing the ABLLS
• Early Learner ABLLS RFT
• Improving EL ABLLS with RFT
• Advanced Learner ABLLS
• Improving AL ABLLS with RFT

171
Assessment of Basic language and Learning
Skills ABLLS
PARTINGTON SUNDBERG (1998)

• Based on the behavior analysis of language
• Identify critical skills to enable a child to be
  a more capable learner
• Attempts to assess function, not just form
• Hierarchy of skills - 25 tracks and 476 items
• Criterion referenced, not standardized

172
ABLLS TASK SET G
173
ABLLS

• ABLLS
• Introducing the ABLLS
• Early Learner ABLLS RFT
• Lipkens, Hayes Hayes (1993)
• Advanced Learner ABLLS
• Improving AL ABLLS with RFT

174
Early Learner Profile Carbone 2004
ABLLS

• Skills between the ABLLS tracks that correlate
  with initial learning performances
• It is recommended these skills be mastered before
  attempting higher numbered items
• Skills include - Visual (B track), Receptive (C),
  Imitation (D), Echoic (E), Mand (F), Tact (G)

175

• B
• Visual Performance
• B1-B5

• E
• Echoic
• E1-E4

• F
• Mands
• F2-F5

• C
• Receptive
• C10-C15

• D
• Imitation
• D1-D7

• G
• Tacts
• G1-G5

176
Non-Arbitrary
Arbitrary

• B
• Visual Performance
• B1-B5
• D
• Imitation
• D1-D7
• E
• Echoic
• E1-E4

• C
• Receptive
• C10-C15
• F
• Mands
• F2-F5
• G
• Tacts
• G1-G5

177
What does the arbitrary / non arbitrary
distinction suggest in terms of training?

• 1 -- Non-arbitrary tasks primary?
• 2 -- Non-arbitrary tasks easier to train?
• 3 -- Non-arbitrary performance probably
  facilitates arbitrary performance?

178
Specific vs. Generalized

• B
• Visual Performance
• B1-B5
• D
• Imitation
• D1-D7
• E
• Echoic
• E1-E4

• C
• Receptive
• C10-C15
• F
• Mands
• F2-F5
• G
• Tacts
• G1-G5

179
Imitation
TRAIN IMITATIVE RESPONSE
TRAIN IMITATIVE RESPONSE 2 TRAIN IMITATIVE
RESPONSE 3 TRAIN IMITATIVE RESPONSE 4
MET
TEST FOR NOVEL IMITATIVE RESPONSE
180
ABLLS

• ABLLS
• Introducing the ABLLS
• Early Learner ABLLS
• Improving EL ABLLS with RFT
• Advanced Learner ABLLS
• Improving AL ABLLS with RFT

181
Develop Exemplar Pool for Skill
Select FS Array
Randomly Select Exemplars
Conduct 1st Training Set Yes / No
Fail
Train Items to Criterion
Pass
Determine Skill Mastery
No
182
Non-Arbitrary
Arbitrary

• B
• Visual Performance
• B1-B5
• D
• Imitation
• D1-D7
• E
• Echoic
• E1-E4

• C
• Receptive
• C10-C15
• F
• Mands
• F2-F5
• G
• Tacts
• G1-G5

183
Matching-To-Sample
B
Visual Performance

• Objects to Objects ----------------------------
  B1
• Pictures to Pictures ---------------------------
  B2
• Pictures to Objects ---------------------------
  B3
• Objects to Pictures ---------------------------
  B4
• Sorts non-identical items -------------------- B5
• Note basic category dogs, trees

184
ABLLS Performance Criteria

• B1 10 objects in FS of 3
• B2